The following response comes from Fred Petty, the director of the Mental Health Clinic at the Dallas Veterans Administration Medical Center and profesor of psychiatry at the University of Texas Southwestern Medical School:

"The current best estimate is that adult lifetime prevalence of bipolar disorder in the U.S. is about 1 percent for severe cases and 2 to 3 percent if milder cases are included--cases that would include cyclothymia and 'bipolar 2.' Given the population of the U.S. (about 250 million), those numbers translate into about two to three million with severe bipolar disorder and some five to seven million total, including those with mild symptoms. So the figures you heard on the radio seem reasonable.

"I have seen estimates that two thirds to three quarters of the people with depression are undiagnosed and untreated. I haven't seen the corresponding statistics for bipolar disorder, but I would say that, looking at mood disorders overall, most patients are still undiagnosed and untreated.

"Right now the diagnosis of bipolar disorder is a clinical diagnosis, which means it is made by a physician based on information about the patient's symptoms, mental status and medical history. It is important that people obtain a medical evaluation, because there are some conditions, such as thyroid disease, which can mimic the symptoms of bipolar disorder.

"There is, at present, no routinely available biochemical diagnostic tests for bipolar disorder. My laboratory has published some work looking at the blood levels of an amino acid called GABA (gamma amino butyric acid), which functions as a neurotransmitter in the brain. Bipolar patients often show low levels of GABA in the bloodstream. The GABA test is currently used only for research purposes; GABA levels appear abnormal only in about one third of the patients who are bipolar. Also, this test is not specific for bipolar, because low GABA levels are also found in people exhibiting non- bipolar depression.

"One of the likely reasons that GABA is not a universal indicator is that bipolar disorder is heterogeneous. In other words, the syndrome we call 'bipolar' actually results from several different causal conditions (etiologies); it is possible that in some cases the low GABA level is the predominant etiology, but not in all. In this sense, bipolar is somewhat like diarrhea--we call it one thing, but you can get it a variety of ways (through colitis, food poisoning, and so on)--or like heart failure, which has a multiple etiology. But low GABA readings look useful because the anomaly persists even after the person improves clinically. If my group can get its grants renewed, we would like to look at GABA levels in children whose parents have bipolar disorder.

"There is also some very interesting work being done with PET scans, which can show deficits of the neurotransmitter serotonin in the brain, associated with depression. Serotonin deficits seem to persist even after the depression lifts--suggesting that it is a real biological marker. PET scans of serotonin in bipolar patients have not yet been done, to my knowledge. Brain imaging using magnetic resonance has been applied to bipolar research; some bipolar patients show brain lesions in these images.

"Another very exciting area of investigation involves the search for a genetic locus for bipolar illness. Bipolar is probably the most genetic of all psychiatric illnesses, and psychiatric illnesses in general have a strong genetic component. If you have one parent with the illness, you face a 25 percent risk; if both parents are bipolar, you face a 50 percent risk. Among identical twins, there is an 80 percent concordance for bipolar disorder. Ideally, we would like to diagnose the illness in young people before they express any severe symptoms. In this way, we could save a lot of human suffering; the earlier a patient gets into treatment, the better the prognosis.